The Airline Went Bankrupt, so Google Bought The Data

Bren: Hello and welcome back to This Is Not A Data podcast. Today we're digging into a story about Google and the purchase of an airline. Turns out this is not for the planes or routes, but for its data. Years of emails, Teams messages, spreadsheets, and even HR records, all of it heading off to train AI models. So we asked the question, if a company's internal data can be sold off like any other asset, what happens to everything you've ever written? on a work account or shared with a business that you trust. And we pose the question, could you one day get paid for it ourselves? Now stick with us because this one raises more questions than it answers, but let's get into it. Hello and welcome back to another, this is not a data podcast and as ever, I mean hopefully if you've been listening by now you know what the answer is but I have my very good friend, Leno with me. How are we doing?

Lenno: Leno, yeah I'm doing well.

Bren: Love it. Love it.

Lenno: It's done and behind us with the nice weather over here in the Netherlands.

Bren: You

Lenno: I hope you are still enjoying the overwhelming sunshine over there in London.

Bren: Well you... so... I tell you what, know how to kick a man when they're down, don't you? Because we've had... I mean, A, we've had emergency texts across the UK because the government's telling us it's hot. mean,

Lenno: Okay.

Bren: emergency blasts to say, careful, it's hot outside. Good call. That was very much needed. No, no... No.

Lenno: You didn't notice, right? No, okay.

Bren: everywhere sold out of portable air conditioning units. I wish we'd started an AC business when we started this podcast. I know we'd be sat on the beach somewhere now.

Lenno: Yep. Yep.

Bren: But, and this is the funny thing, I'm a huge cricket fan, big cricket traditionalist, so I like the long form five day test format. It starts today, which is great. England are playing Pakistan and the reason, and so it's Wednesday today by the way, for the day we're recording. Now the irony is, you can be guaranteed that when the cricket starts it's going to rain. And... guess what?

Lenno: Here we are. It's raining, Okay. But then that's also why when it starts to rain, it lasts forever. Because that is

Bren: This is it.

Lenno: similar. That's my parallel towards cricket. I don't know cricket too much, but I've understood that a cricket game can last for a long period.

Bren: Why are you so mean to all us cricket fans? That's so unfair.

Lenno: I have very, very, very narrow experience or exposure to cricket. So I can't relate actually to the game in itself. This is the one fraction of information that I have. So my context, please widen my world of experience here.

Bren: Well, I mean, so I mean, the best way to describe it is you can play for five days and end up in a draw. And that's something that baffles many. But without going down sort of cricket memories, some of the best and most exciting series have ended up taking five days. And The Ashes with England, Australia and that sort of thing.

Lenno: Mm-hmm.

Bren: But the irony behind having had we've got hosepipe bands We've had lots of sunshine, it's been swelteringly hot. And then the day that the cricket starts, out comes the rain. I mean, that's reassuringly British. think if anything, this is, I mean, it gives us also a great intro for our podcast, hey? So, gotta embrace it.

Lenno: we should and we need to. Yeah.

Bren: But we haven't come here to talk about the weather, although maybe that is this is not a weather podcast that we can start on the on the side. Yeah,

Lenno: We can start a spin-off,

Bren: I'm expecting you to hold up the hand to be like across the the west of the country. This is what it's going to be like. But you you sent a link to an article that I think is really worth us talking about today because there's some quite. big activity when it comes to data and the business world. And because of the name of our podcast, because it's not though this is a weather podcast, it's all about not being about data. Google seemed to be on the move in acquisitions, which is not necessarily a new thing. But what struck me was the reason as to why they bought the business. Do you want to for people listening, do you want to? Just give a bit of a synopsis of the article that you sent.

Lenno: Yeah, I think you're teasing layup is excellent. so kudos, kudos when kudos

Bren: We're getting better at this, aren't we?

Lenno: are due. No, indeed, I came across an article yesterday stating that Google actually bought a portion of a bankrupt airline. So the airline in itself was hit hard during the COVID pandemic

Bren: Okay.

Lenno: and declared bankruptcy. And now... Google wants to train its AI models better. All of the AI models to date, most of it have been trained with publicly available data. And based on that, the models have been shaped and have been educated. Now, the thought behind this is the purchase of a company's data. And this company's data contains text messages emails, Teams messages, spreadsheets, engagements that not only internally happens, but equally also customers and suppliers support partners of that airline have been engaging with the airline company in itself. So there's a big, big data set that has been acquired. from this bankrupt estate. And of course when a company goes bankrupt there's assets that one can buy. a company's valuation and we've been talking about that, I think for years, whether data is an asset. asset's data has been reflected as an asset a lot. However, is never put, or at least not in the companies that I've seen so far, has never been put on the balance sheet as an asset. Which from a financial perspective, it's a little bit painful. financial perspective and a data perspective a little bit painful because we do say we need to treat it as an asset, but it's not treated as a genuine physical or software related asset. Nevertheless, that's beside the point. Now, apparently, there is value associated with the company's data. And we're not even talking about the creation of your data to drive your company to exist. This is genuine interaction data that has happened. You mentioned before we started this podcast also, that as part of due diligence and as part of acquiring company, there's of course, also evaluation of all of the information that is available and everything goes from one organization to the other. In the early days and yeah, sorry for me, the early days goes back like 30, 20, 20, 30 years or so. That also was part of the estate that a company was acquiring. However, we did not have the technology. We do not have the means to sift through all of that. that data, all of the emails, all of the messaging that occurred, all of the spreadsheets, all of the Word documents, all of the presentations, et cetera. You would not have the manpower to go through all of that information till now. Because nowadays we seem to have models that are capable of running 24 hours a week, sifting through all of that information and then come with summaries, regeneration of direction, maybe even new business models based on all of the information that is available. The vast amount of information that you can distill from this. And so my point being, I was triggered by the article in itself. Second, there is monetized value now with your data. And the data in itself can be very, very big. It is going to be trained and that is also part of that article. It is going to be supporting the training of the AI models that Google is using. Now that brings a couple of points to mind as well, because my question around it is, do I now need to start paying even more attention to what I'm sharing with a company as a consumer, but equally also as an employee within a company? Of course, you need to be politically savvy and... be able to interact with a lot of people in different mentalities and frames of reference. Which is more, I think, more exposed when you are engaging in groups. But equally now, in the one-on-one conversations where maybe you can be harsher. because of the one-on-one engagement, because of the depth of the conversation that you're having. To what extent everything is going to be recorded or going to be, is everything being recorded? We're recording this podcast and that goes into the public domain. I'm aware of that. You know, that's

Bren: Mm-hmm. Yeah.

Lenno: fine. The one-on-one conversation that we're having outside of the recording. is in my mind, that's outside of the recording. Now, if we are having that conversation, and I'm positioning this like to have your reflection to it. If we were having a WhatsApp call. which is part of the Microsoft estate.

Bren: Mm-hmm.

Lenno: And, no, sorry, what is it? The Meta. The Meta.

Bren: Yeah.

Lenno: And Meta decides to sell off their data. Is our conversation proprietary? Is our conversation public? Is our conversation code to be treated into that modeling training? I may be ignorant in this space. What is the frame of privacy in the information that is generated on the basis of our conversation, the emails that I'm sending, the Teams messages internally and externally? Apparently there's value now in buying all that stuff.

Bren: It's a bit mind-blowing, it? Really.

Lenno: It is-

Bren: mean, just the great matter is actually going like the clapboards at the moment because I think the couple of things come to mind when you join and this is very much for the corporate world and you've probably seen it in your career as well. When you join a company, I'm going to say private company because I've not worked in the public sector, you know that every message you send, email, Slack, Teams, whatever the corporate messaging platform is, that is being recorded. Many at the beginning forgot that, so would therefore, know, slater colleague or say something that they wouldn't necessarily say to sort of out loud. But that's fine, because when you join, you acknowledge that you are part of somebody else's environment and everything that you've sent is saved. I remember my training at EY very much put me at a good starting point because they said, everything is recorded. I mean, it was back in Lotus Notes back in the old days. But every message, every communication that happens on corporate devices is going to be captured. So you go like, okay, cool. Which I don't necessarily think is a bad thing because you need to know that people are doing the right things. Where it gets a bit messy, which is what you're saying, when you leave or if a business changes, What happens to that information? Because when you were talking, the first part that comes to mind, and you made reference to it before we started recording, is what about GDPR? If we live in, I mean, you live in the Netherlands and I live in the UK, and if either of us had had any interactions with spirit, I think it was, that was born, could we now submit a data subject access request to Google to understand all the information they've got on us and then allow them or... prevent them from using our information. I don't know, but this goes into the realms of, you talk about monetization, I mean, that is the holy grail of where I guess the data world wants to go. I I've had many conversations around how do you value people on the balance sheet? Because I think that is also an aspiration

Lenno: Mm-hmm.

Bren: that many want to try and capture. But from a data perspective, For Google, it's really smart. mean, you're to get the warts and all conversations, the genuine comms and how people act and interact and what they've

Lenno: Yep.

Bren: said. I mean, that is an absolute treasure trove of goldmine of true human interaction.

Lenno: Now, think on the CNN publication that I have in front of me, it does state that the data has been stripped of anything that would allow individuals to be identified according to the court filing.

Bren: Do you believe them?

Lenno: Nevertheless, there is emails, internal communications, spreadsheets, transactions with the public, including bookings and frequent flyer information, as well as human resource information on its employees. It's a massive, massive amount of information that is being monetized now.

Bren: Yeah, but I think, and this might be a slight counter and feel free to disagree. In previous places of work, if you've had a partner or an executive who's been with the business, let's say 25, 30 years. So a vast amount of knowledge is held by an individual, probably captured in emails and notes on machines. When they leave or depart the business, Having access to that, think is actually quite, could be quite valuable to understand what happened in a deal or a transaction or an event 17 years ago. I've been able to, I don't know, take the PST file from an inbox and say, you know what, in this scenario, what did we do? Or summarize this. I think there's an element of that could be quite useful, but it verges on. sort of in interruption of privacy and all that side of things. But as a private company, you have that choice to do or not to do and then go through the ramifications. If we flown with spirit, we have zero. We are now just at the mercy of being part of Google's Gemini and whatever it is. And that doesn't sit well.

Lenno: I do agree with that assessment. I also recognize that it would be extremely powerful to somehow retain experience from

Bren: Hmm.

Lenno: people that need step out of an organization. How do you do that? I think a of years ago. We did a small workshop with, I think it was Cambridge University, on how experience of people could actually be sort of documented. And by

Bren: Okay.

Lenno: that, allowing us to move forward with that baseline experience already.

Bren: Hmm.

Lenno: So I definitely see the possibilities, the value of it, and the ability to move forward. I think it's in this publication as well, it's essential that identifiable information is taken out of the entire data set.

Bren: Great.

Lenno: I personally, I'd be interested to understand the legal framework into more depth around this particular example. Because to what extent is there personal identifiable data a factor when it is core company information? I definitely understand the the frame of reference when I'm a consumer and I'm engaging with an airliner. And then as a consumer, as a private individual, you're much more protected from a data standpoint. So having that type of identifiable information out there or out of it, I can imagine, but take the metadata on your PowerPoint documents that you've generated somewhere in 2008. You know? Your name is connected in there. Are we stripping out that particular metadata? Are we stripping out maybe references that I've made in that document in itself to individuals? Is that considered to be personal identifiable data? I would argue it is.

Bren: Agreed.

Lenno: Nevertheless, stripping out, and that is what the court filing says. Identifiable data has been stripped out. To what extent and to which depth that actually has occurred across the millions of records that have been shared. Who will be able to identify it? The only one that is probably able to identify it is an AI in itself. Because you and I are not going to go sifting through all that documentation.

Bren: No.

Lenno: Or maybe you want to do that in your spare time when you're in a tent somewhere.

Bren: I mean, I think you're absolutely spot on because the headline parts are going to be, you know, the PII, anything that is contentious, being able to tie it back to individuals. We have, as consumers, we definitely have or should have the acknowledgement that everything that we send or put out into the ether is captured and recorded and it can be played back to you. That should be the starting point, but I think we often forget that. certainly if you're thinking about the chatbot conversation we had previously, Leno, if you go back and you ask everybody that's interacted with an AI chatbot, if they wanted to see all the messages that have ever been sent by them, I reckon you're gonna get some pretty embarrassing, embarrassing responses to that. not that I'm, unless we start, this is not a politics podcast. you know, from a political standpoint, often messages that shouldn't become public become public. And then there's a bit of a hullabaloo and a kerfuffle around it. I think is the actual lesson is it a people lesson to this of do we need to be more conscious of everything we do is going to be made public. But to count on my own point, if you have to sanitize and really think about everything that you're going to send, how are you actually going to get to the answer that you you need? It's going to be no curiosity, can be no debate, mean debate is being lost anyway, you know, are we just on a one-way street to a world of bland sort of arbitrary sort of data nonsense?

Lenno: Welcome to the matrix. And now if,

Bren: Hahaha!

Lenno: even where everything is going to be generated in a digital world that is based on rules, I think it's the likelihood is high that ultimately we're getting towards exactly the same answers over and over again. So is that going to be more of a blend vanilla?

Bren: I'm

Lenno: Maybe, probably. To your previous point, I do recognize or think that we are quite ignorant to the fact that all of our data we're already sharing. Everybody with a smartphone, a smartwatch, a smart whatever that it may be, you will have your location data on for certain applications or it's highly convenient. Are you doing any type of navigation? Yeah, from one place to another, I will go running. I'll do tracking of the routes that I'm doing. I will have documentation of the pace, heart rate, et cetera. So there's a whole lot of convenience that is coming along with that. Nevertheless, I'm storing that on the device or in an application that is out there in the public domain. So I'm already sharing my data. I think more and more I think that the legislation around for instance GDPR is mostly there to be conscious about what is being done with that data. can I assume or expect as an individual that all of my data is protected from the moment that it is generated all the way down to whenever, I don't think that you can. I don't think that you can, and I don't think that it is realistic nowadays to also expect that that is the case.

Bren: What about... So the whole concept of buying Google buy in this business was access to data and maybe unpopularly from a pure business perspective, I think it's a really shrewd move. I can see exactly why they've done it. But in the same premise... Meta is also in the press at the moment because there's a number of litigations happening in the states around social media and I'm not suggesting we go down that route

Lenno: Yeah.

Bren: but if one part that did come to mind and maybe this is a slight public naivety if you're not sure what the product is you are the product and we should we should never forget that but kind of pulling on that thread just a little bit more if Google are buying businesses for access to underlying data that's created by users. Can we get to a stage or will we get to a stage whereby we allow companies like Google to monetize our data in return for a fee? If you did a GDPR

Lenno: Ooh.

Bren: request to Google and said, is all the information and asked for everything you've got.

Lenno: Mm-hmm.

Bren: Okay, fine. I'll allow you to use it, but you need to pay me X a month or Y one off. Is that what we're going to get to?

Lenno: We can, we, we all that that's an interesting business idea. We can flip the whole table. We can flip the whole, the whole setup, the business model. I can now you can subscribe

Bren: Because if you think about it,

Lenno: to my data.

Bren: yeah. Because think about all of the information that you've put out into the public domain.

Lenno: Yeah. Yeah.

Bren: And I'm reasonably laissez faire about it. And I know my cyber friends will be pulling their hair out at this point. at some point, will it flip? so we, know, Leno's profile, Garmin want to be, okay, we know that you go trekking. We know that you do this. We'd like to buy your data from you at, I don't know, $10 a month. for the period of whatever.

Lenno: Mm-hmm.

Bren: Okay, cool. Well, you're going to capture anyway, does that work? And then you could do the same with Netflix, so they'd be able to monitor your viewing

Lenno: Yeah.

Bren: habits. Or I mean, I haven't really thought this through in great detail. But is there a is there a tipping point? Where that we are all the products by virtue of that we give all this information away for free. They're making bazillions of money, pounds and dollars off us. And we talk about like AI being a bit of a black box and just,

Lenno: Mm-hmm.

Bren: you know, can it be trusted? Is it safe? Is it being trained on the right things? For those listening, I put all those in air quotes. It's like, well, I'm not sure. I fear or wonder whether there'll be a little bit of a shift, but it's gonna take a fairly monumental act. that to happen and maybe

Lenno: Mm-hmm.

Bren: the meta conversations are or the Google one. I don't know maybe that's too much of a rabbit hole but what I mean what do you think?

Lenno: But a rabbit hole it is. A

Bren: Hahaha

Lenno: rabbit hole it is. I think we can go into all sorts of directions and depth on that one. Our platforms, all of the platforms that we're using, one have momentum, two have a whole lot of subscribers and people leveraging the platforms as well. And because of that, indeed, we as subscribers are the product in itself. Initially started off with a high level service that we're getting for basically zero investment. Except for giving away a portion or all of my information. And that is like physical products that is now being tuned down into a much more push of information towards us, AKA commercials. So we are becoming the products and the recipient of what is it, sales tunnels from these platforms. And you've named a few, so Meta, but Google, indeed, and I think Spotify, Netflix, well, you name it. There's all kind of commercials that is coming towards us. And of course you can then pay a fee to get rid of that to a certain extent. indeed we are getting the, we are being the product. Now

Bren: Hmm.

Lenno: to flip that table, I think that that's where the tricky point is. Us individuals somehow need to group. to a significantly large scale to actually make a dent and make an impact. And I not necessarily see that happen directly.

Bren: I think you're right. would take a wave and a really... It will either take a lot of people. It'll be a volume or scale piece. It will either take a monster

Lenno: Yeah, exactly.

Bren: event like the meta legislation or the lawsuits to change or

Lenno: Yeah.

Bren: a wave of people. But I do wonder... And it's slightly tangential and it's a conversation I had last week. I used to wear a whoop, one of the fitness...

Lenno: Yeah,

Bren: And very cool company, terrible name in my opinion. And I mean, I was fortunate enough to listen to Kristen Holmes talk, who's the she's effectively the principal data scientist or data chief data person at the business. Not long ago, very, very impressive woman, fascinated to listen to and talked about sleep and so forth. Now, I wore one every day for 18 months. This is probably four or five years ago.

Lenno: Mm-hmm.

Bren: It's a subscription model. So you spend 25 quid a month, whatever it is, to wear this thing that tracks all your information and tells you you've had a really crappy sleep, even when you've slept really well. And when you don't feel like you've got anything else to give, it'll tell you you should push more and you've got more in you. So we'll leave that there. But what was interesting, and I said this to somebody I met last week, the moment you stop paying, you lose access to your data.

Lenno: It's gone.

Bren: Well, it's gone to you. So I ended up submitting a GDPR request to say, can you tell me all the information that you've got on me? Because I want a copy of everything that you've tracked over every day for the last 18 months. And I got the feeling at that point, it was a bit of an unknown because they they'd not really had that. But if

Lenno: Mm-hmm.

Bren: and maybe it's sort of my geeky data route. But on the principle that we're paying, I mean, Google will probably do something with it. But the moment we stop using a service, we presume that that data just automatically disappears or it's, you know, it's what it's been done for. And I think this is probably the naivety and where we need to be really cognizant. What happens with that information? Even if we don't know what happens, we should know that something will. And that I think is where we've got it wrong in the past. Would I want my... personal WhatsApp shared publicly? No, because it would just be embarrassing. I mean, it would also be really boring to read. those

Lenno: Yeah.

Bren: messages you write on the presumption that they're not going to be made public. And I think the work that everybody would have done at Spirit would have been done on the presumption that it won't be made public or used publicly. Whereas now, if I think back to my corporate days, if every data strategy and data insight I wrote, that was very specific for a task that we were doing for the exec or whomever. If I had to ride that on the presumption that at some point it might go live to the public or be used to train a model, would I have written that differently? Maybe, maybe not. But if that's a conversation or a question you have to think about, then that puts a very different complexion on how we as data executives need to manage people and situations and all sorts. I think all it does do is it creates more and more questions rather than answering any of them. But it does make me sit back and think, holy, what's going on?

Lenno: Yeah, 100%. And yeah, unfortunately, this again is raising more questions than giving answers in our

Bren: Ha ha!

Lenno: conversation.

Bren: I tell you what, maybe this is our secret sauce letter. We are just very good at talking nonsense for half an hour and then asking some really powerful questions at the end. Although though in fairness, I'd say actually this has been a very insightful conversation and one that hopefully will instigate a few questions. But maybe that is the answer is that it just creates more questions and we need to be very aware of them.

Lenno: Yeah, continue to have the curiosity and also the resilience to work through it.

Bren: And I mean, thank you for sharing the article at the very beginning, because it was, I mean, fascinating. But if you if you had to kind of summarize your feeling about reading that in a sentence or two as we close out, what? What would you say?

Lenno: think be very, very cautious about conscious about how data effectively is used nowadays.

Bren: Yeah.

Lenno: That is one thing that every individual needs to be cognizant about. But second, data now has a tangible monetizable value. And this is a mantra that we've been talking about for 20, 30 years.

Bren: Hmm.

Lenno: Data. to be treated as an asset. Apparently now there's tangible evidence that data can be treated as an asset.

Bren: tell you what, that's powerful. By the way, to make you forgot, I'm going to have it put as a mantra on the board behind me so that every time we watch, you'll be able to see the world according to Leno's. I'm still trying to push this. yes, we'll have it on the wall behind. But now, you're right. Certainly, I don't think you're wrong. And that's not a way of hedging. I don't think we really know. if we're maintaining... the curiosity is, as I read not so long ago, life is not about knowing the right answers, it's about asking the right questions.

Lenno: Yeah, agreed.

Bren: Well, I mean, 36 minutes in almost this, mean, fascinating and really that the mind is going 10 to the dozen at the moment. I think there's gonna be plenty of conversations we're gonna need to have as we follow on from this. But as ever, for everybody listening, this is the time where we need your input because the, you know, the algorithms, the AI stuff, they listen to this and they respond very well to five-star reviews. So we need to subscribe. I think the YouTube people are like, smash the subscribe button. I don't really know what that means, but basically push the subscribe, give us a five-star review. mean, other reviews are available, but we really like the five-star ones and share the pod amongst friends and colleagues. I mean, I see you smoking it out. mean, one day I'm going to get you to do the outro.

Lenno: Yeah, Yeah. No, I love it. How you summarize this and indeed also the reviews of it.

Bren: Well, that's very kind. But Leno, as ever, thank you and to everybody else. We'll see you next time.

Lenno: Aloha.